RD Tech Web Bounty

[onleavedontatme]

GoodPoster2001 - Today at 12:19 AM
i'll pay $25 USD to the first person to come up with a cool and engaging replacement for R&D and anothr $25 for someoen to code it tbh

oranges - Today at 12:19 AM
I'll pay 200 USD to whoever makes kor's tech web

PM me and we can hash out details if you're interested

[ShadowDimentio]

TO THE FLOWCHARTS!

[batman spin]

[Gun Hog]

Why, Kor? How is this not going to be anything but a huge nerf? Is this something that will make people actually want to play it more? I am sorry to be so biased, but I have never seen any proposed rework be anything other than something that makes R&D take a lot longer to do, be a lot more expensive, and/or remove access to top-end equipment. With this being financially motivated, I am now genuinely afraid that I am never going to see a bag of holding or durand ever again. Whatever you guys do, please do not make all the high tech stuff as rare as Phazons.

I am hoping for something interesting where Science can kit out other departments with ease (lack of viable distribution methods is a big flaw right now), and can still be as rewarding as it is now, but since I have not lately seen anyone other than Kev try to do anything other than a nerf, I hope you can understand how I would be worried about this.

[ShadowDimentio]

Here's the flowchart I made, any R&D mains can come to suggest how things should be moved around but I think it's pretty good.

[Bawhoppennn]

Me and Kevinz already had this figured out before this post, this is a lucky break for me

[kevinz000]

right the question is how the fuck do i code a fucking tgui tech web until i have that out don't expect anything.

[Remie Richards]

>asking help
>to receive $200

Haha, nice try kevinz, I doubt you'll get much out of anyone for a while.

[InsaneHyena]

Why, Kor? How is this not going to be anything but a huge nerf?

Good. R&D deserves more than a few hits with a nerfbat. Enough is enough.

[onleavedontatme]

@GunHog

It might be a nerf to the top rnd players but I think that will be balanced out by other people being able to do rnd without an external guide. Also being able to go do xenobio/experimentor/toxins while the machine researches instead of fiddling with the shopping list.

[Steelpoint]

A tech web strikes me as but one part of a larger overhaul to how RnD is performed. I won't speculate on how a new RnD system would work but I'll be interested to see the future proposals.

[kevinz000]

Why, Kor? How is this not going to be anything but a huge nerf?

Good. R&D deserves more than a few hits with a nerfbat. Enough is enough.

we asked you.

[D&B]

>RnD powershitters crying

Delicious

[ohnopigeons]

A tech web is code complex and doesn't solve the issue of powergaming.

Instead of a tech web do stellaris' random tech deck, this will force scientists to strategize their research on the fly instead of following a guide on the most optimal path.

Also tech webs are gameplay complex and don't do anything except make research even harder and more unfun for new players who still don't understand the current RnD systems while top rnd players will be annoyed but still be able to powergame effectively to their hearts desire.

[bman]

this is my proposal
https://tgstation13.org/phpBB/viewtopic ... 81#p310981

[captain sawrge]

eye'd like to note the bounty will be delivered to the Ideas Guy and the coder upon the whole thing being merged.
If you code your own idea you just get the whole thing I guess. First person to finish something and have it merged gets the money

[bman]

eye'd like to note the bounty will be delivered to the Ideas Guy and the coder upon the whole thing being merged.
If you code your own idea you just get the whole thing I guess. First person to finish something and have it merged gets the money

reeeeeee he changed the goalposts triggered

very well then

[captain sawrge]

eye'd like to note the bounty will be delivered to the Ideas Guy and the coder upon the whole thing being merged.
If you code your own idea you just get the whole thing I guess. First person to finish something and have it merged gets the money

reeeeeee he changed the goalposts triggered

very well then

Ideas don't mean a whole lot if they're not getting coded.

[bman]

eye'd like to note the bounty will be delivered to the Ideas Guy and the coder upon the whole thing being merged.
If you code your own idea you just get the whole thing I guess. First person to finish something and have it merged gets the money

reeeeeee he changed the goalposts triggered

very well then

Ideas don't mean a whole lot if they're not getting coded.

yeah but kevinz is already coding the shitty tech web for his whore money

[onleavedontatme]

eye'd like to note the bounty will be delivered to the Ideas Guy and the coder upon the whole thing being merged.
If you code your own idea you just get the whole thing I guess. First person to finish something and have it merged gets the money

reeeeeee he changed the goalposts triggered

very well then

Ideas don't mean a whole lot if they're not getting coded.

yeah but kevinz is already coding the shitty tech web for his whore money

Scared?

[onleavedontatme]

Instead of a tech web do stellaris' random tech deck

Please elaborate on this and how it might apply to SS13 it sounds interesting

[bman]

eye'd like to note the bounty will be delivered to the Ideas Guy and the coder upon the whole thing being merged.
If you code your own idea you just get the whole thing I guess. First person to finish something and have it merged gets the money

reeeeeee he changed the goalposts triggered

very well then

Ideas don't mean a whole lot if they're not getting coded.

yeah but kevinz is already coding the shitty tech web for his whore money

Scared?

no sir im a god fearing citizen and i dont take drug money

[bman]

tbh im not really in this for the money i just dont want rnd to be another tech web

if my idea gets coded just give the other 25 dollar to the coder too

[lumipharon]

Instead of a tech web do stellaris' random tech deck

Please elaborate on this and how it might apply to SS13 it sounds interesting

If you haven't played it, or other games like it, basically research is like this in stellaris:

• You have 3 research paths, physics, engineering and bio or some gay greenie shit.

• Each path can only research one thing at a time (unlocking throbbing red laser blasters, tastier grox meat or whatever).

• You typically have 3 research options for each path available at any one time ('cards' in the deck analogy). You choose one of the three, wait for it to complete, then the deck 'reshuffles' and you get 3 new options (can include the 2 you didn't choose previously, it's random draw).

• Each research 'card' may or may not have certain requirements, usually research prerequisites, but sometimes other shit, before it can even be drawn as an option to research. (for example, must have already researched throbbing red laser blasters to be able to even get the possibility of drawing the pulsating laser rod card)

• Additionally, each 'card' has a category (lasers, construction, particle physics or whatever).

• You can assign a science nerd to each research path, and if they have a preference for certain categories, then research options for that category are more likely to appear (some won't appear at all without the right specialisation). For example, a laser specialist dude is more likely to get the option to research laser related shit.

You could translate this to ss13 fairly well: In the basic you would have one (or more, maybe you could research/build something to unlock more slots?) research slots available. You get a few initial options to choose researching initially, and start working from there. Instead of scientists with specialisations, you instead repurpose the deconstructive analyser, into a non destructive analyser (or could keep it destructive, I unno) so say if you wanted to focus on weapon research, you put a gun in and the next time the deck shuffles you're more likely to get gun related research options. If you want to research bluespace shit, you chuck a bluespace crystal in there and see what you roll.

Some research options can only be rolled if you have high enough prerequisites AND the correct item (or item from a particular small group) in the analyser machine. For a random example, to get the ebow research option, you would need certain research done in energy weapons or some shit, AND a mini ebow in the analyser.

So research options can be broadly placed into two categories:

• Options that actually unlock one or more items to build
  Options that are purely knowledge/understanding, which only serve as prereqs for other research options.

So maybe you to get pico manipulators you would have had to already researched nano manipulators, and also 'advanced subatomic assembly' or whatever other fluff shit, with a higher chance of being drawn if you have a nano manipulator in there, because it's a lower tier version of the same thing.

[Jordie0608]

That does sound more interesting.

[captain sawrge]

It's ultimately up to Kor but I really hate the destructive analyzer and the former $25 was basically just for anyone to come up with something more interesting or engaging than it that is viable to code.

[Steelpoint]

I should also add that the Stellaris research system has weighting behind every card, and some cards are guaranteed to be on offer early on.

Essentially cards can be divided into rarity, with rarer cards having a less common chance to appear. In addition cards in the same category have different weights so they have a higher or lower chance to appear.

[lumipharon]

In my proposal it's just an 'analyser', which makes it somewhat more likely to be to get related research options rolled, but all the other requirements are still needed - in essence it's just pushing the rng in a direction of your choice (although as said, certain options NEEDING a very particular/unusual item to be analysed I reckon is neat).

Or to put it in stellaris terms, you could have a science bro specialised in the terraforming research, and still get precisely 0 terraforming options fucking ever, and have 10 worlds you can't even colonise/tile clear.

[kevinz000]

welp i just threw out origin tech code.
honk. next on my list is completing node viewing for rnd console and some points var to store it.
As for the actual decon/points system, that sounds really interesting, i'll look into it a bit and see if kor will agree to that.
Probably better than passive gen anyways but w/e.

edit: it also sounds like a pain in the ass to code, but if we want it I can make it happen.

[kevinz000]

Oh yeah I ran into a problem with the points system, as to where the points are stored. Obviously it can't be generated per RD console or R&D server, as you could build shitloads of them to cheese the system. Setting it globally allows for some griefer in some remote corner of maint to sink it on random crap. Ideas?

[Iatots]

>increasing processing power should not increase research data output
dumb catperson

[kevinz000]

>increasing processing power should not increase research data output
dumb catperson

the problem will become "oh wow a powergamer built 100 servers, now rnd is going at 100x the speed as usual!"

KOR, PLEASE ADVISE!

[Remie Richards]

exponentially worse gains. simple.
1 = 1 point
2 = 2
3 = 2.5
4 = 3
5 = 3.5
6 = 4
7 = 4.25
8 = 4.5

etc.
value in doing it without it snowballing.

[Iatots]

Make servers produce heat and drain lots of energy. Mining isn't cheap.

[kevinz000]

Right, I'll make servers take more energy to run and produce way more heat, but each would generate points based on how many there are that are currently generating points. However, there's still the problem of "What if someone builds servers in places with no RND connection and blocks access to sabotage science?" Best I can do without any further ideas is have them output points to a global variable that console can take from, but that still raises the issue of "Someone builds a research console in some remote corner of the station and sabotages science by draining points without actually syncing" so they can research the same thing, and then destroy the research, and waste all the points.

Actually, what if research consoles no longer did the research, but instead, research servers are controlled by consoles to give research points into one node to automatically research it? That way it's much harder to grief without a counter, as all servers will sync with each other.

[Iatots]

Ok I need to get a couple things straight:

You want the "stellaris" system, correct? "pool of random 3, pick one then reshuffle".
The research process involves selecting a specific technology and having said tech researched over time, not simply by its passage but by generating "research points".
The problem is in how these points are generated: if consoles generate them, multiple consoles could speed research up, or even research different paths of the tree, defeating the point. If servers generate them, well the problem is pretty much the same.

Since I think arbitrary restrictions on numbers of consoles or servers is dumb, I propose power as the bottleneck.
Computers take power to run, and a lot of power to research with.

Now the problem that remains is: how do we treat multiple consoles and server, in good and bad faith?

Here is my proposal:
All button pressing is done on a console with the research program. This console links up to a server, where the tech tree is stored, via a network. The research points are generated by computation power from all consoles on a server's network running the research program. You couldhave it be a modular console thing, since processors are a thing, but one step at a time here. The safety of the network could be as simple as requiring a password before connecting from any console, with network setup done from the server computer. Mass building consoles could forseebly act as a powersink of sorts, but I believe engineering can now work on APCs remotely, so as long as they can notice the obscene power drain they should be able to deal with it.

So consoles make the points, consuming power, and all points go to a server. Said server also holds the unlocked tech tree and is secured by passwords and shit. Multiple consoles researching different techs may be an issue depending on how good of a bottleneck power consumption can be.

Network merging, server synchronization and all that fun stuff I leave to someone else to figure out.

[onleavedontatme]

Power is meaningless though, the SM generates like 10x what you need to run the station and thats before you set up the turbine/pacmans/solars

[Iatots]

Golly gee it's almost as if every department is a mass of stitched together features in needs of some love.

[kevinz000]

read pr for latest updates

tl;dr research servers gen points with diminishing returns, i'll be upping their power draws and temperature draws considerably.

[Steelpoint]

What PR?

E: nevermind found it: https://github.com/tgstation/tgstation/pull/28683

[Wyzack]

It would be pretty sweet if we could make the power draw sufficient to require an advanced engine set-up to sustain multiple machines for optimized research, we would be effectively killing two birds with one stone here

[kevinz000]

It would be pretty sweet if we could make the power draw sufficient to require an advanced engine set-up to sustain multiple machines for optimized research, we would be effectively killing two birds with one stone here

off topic(from me) but i'd love it if multiple deps had machines that take large power draw/requires special circumstances/conditions to run, like how R&D servers require cooling (And they're about to require a lot more cooling and power). It'd give the engis something to do other than brag with their 30 million watt setups.

Edit: Something that isn't robustin's ERT gear points system.

[Bawhoppennn]

Passive gen is bad, mmkay.
Being required to deconstruct stuff if you want to progress is an interesting dynamic with all sorts of cool possibilities that our current system didn't capitalize on.

[captain sawrge]

Passive gen is bad, mmkay.
Being required to deconstruct stuff if you want to progress is an interesting dynamic with all sorts of cool possibilities that our current system didn't capitalize on.

wrong on all counts!

[Bawhoppennn]

How is not actually doing any of your own work in any way interesting?
If you have to potentially sacrifice something in order to achieve progress on something, that can be a comprehensive system that allows for a wide range of possible situations where you have to make various decisions about what you should deconstruct, and shouldn't, thus forcing you to take a balanced approach in how you advance in the R&D system.

[captain sawrge]

How is not actually doing any of your own work in any way interesting?
If you have to potentially sacrifice something in order to achieve progress on something, that can be a comprehensive system that allows for a wide range of possible situations where you have to make various decisions about what you should deconstruct, and shouldn't, thus forcing you to take a balanced approach in how you advance in the R&D system.

This isn't at all how it works in practice and R&D boils down to a ratkid scavenger hunt for 5 minutes and then a UI clicker for another 10 and then bish bash bosh you're printing out explosion guns and infinity backpacks and magic hardsuits

[Steelpoint]

It ain't even a scavenger hunt. You can get most of RnD done with tools inside of RnD, or items you get from research. Some silver from Mining and you can get the station almost fully upgraded.

With a touch of good minerals (Uranium, Diamond, Gold, etc) and a few other good items here and there (Combat Shotguns for example) you can near max research.

Its honestly not that hard to get high levels of RnD in a round. There's a good reason you often see Assistants, Doctors or Miners doing RnD by themselves, because its piss easy and hardly a job unto itself.

You could make RnD public access and nothing would change.

[Bawhoppennn]

How is not actually doing any of your own work in any way interesting?
If you have to potentially sacrifice something in order to achieve progress on something, that can be a comprehensive system that allows for a wide range of possible situations where you have to make various decisions about what you should deconstruct, and shouldn't, thus forcing you to take a balanced approach in how you advance in the R&D system.

This isn't at all how it works in practice and R&D boils down to a ratkid scavenger hunt for 5 minutes and then a UI clicker for another 10 and then bish bash bosh you're printing out explosion guns and infinity backpacks and magic hardsuits

That's why it needs to be reworked?

[XDTM]

How is not actually doing any of your own work in any way interesting?
If you have to potentially sacrifice something in order to achieve progress on something, that can be a comprehensive system that allows for a wide range of possible situations where you have to make various decisions about what you should deconstruct, and shouldn't, thus forcing you to take a balanced approach in how you advance in the R&D system.

This isn't at all how it works in practice and R&D boils down to a ratkid scavenger hunt for 5 minutes and then a UI clicker for another 10 and then bish bash bosh you're printing out explosion guns and infinity backpacks and magic hardsuits

That's why it needs to be reworked?

Because it's pretty boring, never changes between rounds and balace-wise it prevents science from getting new high-end gear since they'll never have to sacrifice something else to get it, so it becomes a straight buff instead of a new option.

Automatic point generation, with some ways to speed the process up, means that instead of following a list clicking the same things every round you can instead wait for the next upgrade and meanwhile actually do what RnD is meant to do, which is building stuff for yourself and others, which changes depending on the situation and is definitely more fun.

The whole techweb thing should also allow a certain degree of specialization, meaning that you could now code stuff like autosurgeons, telepathy helmets or whatever and put them way up in a branch of the tree so you can't have that AND BoHs AND beam rifles AND portal guns (unless the round is really long).

[FantasticFwoosh]

Meanwhile the experimentor just pales out of existance as science's local assisted suicide by fireball malfunction machine because everybody forgets its limited function of converting strange objects into tech giving doo-dads.

[Gun Hog]

Meanwhile the experimentor just pales out of existance as science's local assisted suicide by fireball malfunction machine because everybody forgets its limited function of converting strange objects into tech giving doo-dads.

Or its actual purpose: A murderous DA that has the chance to let you scan an item twice for the full tech benefit, scan it once to destroy the item but still give you tech, or destroy the item and give you nothing. Say I kill an Op for his super high illegal and combat tech weapon. Since nothing Science makes using the tech is as awesome as the gun itself, I can pop it in the E.X.P.E.R.I-Mentor and then try to scan it for those sweet tech levels, but KEEP the gun. The risk I am taking is of course that the machine might eat my gun and try to kill me at the same time. I can mitigate that risk somewhat by wearing one of the suits in the lab that matches my experiment type and also upgrading the machine. Even then, I still might be blown up or middle fingered, but HEY, I enjoy taking the chance.